Hi Yang
That is a good way to do it. Certainly if the surface registration fails the parcellation will as well. There are commands to directly visualize the target and registered source sulc and curv, but they are a bit arcane and I think checking the parcellation is probably sufficient.
Cheers Bruce
From: Alex free_learner@foxmail.com Sent: Saturday, July 18, 2026 11:51 PM To: freesurfer freesurfer@nmr.mgh.harvard.edu Subject: [Freesurfer] How to check surface registration accuracy
External Email - Use Caution Dear experts,
I have a basic question regarding quality control for surface-based registration. In volume-based analysis, I usually check registration by overlaying the transformed image on the target image. What is the recommended way to do this for surface-based registration?
Currently, my practice is to visually inspect the Desikan-Killiany parcellation overlay on the native surface, under the assumption that a reasonable parcellation result would indicate satisfactory registration. I chose the DK atlas because its gyral-based labels are relatively easy to recognize on the native surface. Is this practice sufficient, or are there better practices I should follow?
Thank you for your time!
Yang Hu